A novel way to massively improve the SQ of computer audio streaming

[austinpop]

7 minutes ago, BigAlMc said:

 

Hi @austinpop

 

Quick question. If buying the above from SOTM (SU-1 excluded obviously) and asking them to mod a switch as part of that. Would buying an SMS-200ultra still make more sense than buying a cheaper SMS-200 and asking/paying them to mod it for the sCLK-EX?

 

Alan, I know it's a bit confusing. You need at least one Ultra component, because it embeds the sCLK-EX board. You then need the other component to be be modded to accept the sCLK-EX input.

 

So for your example, you need:

• sMS-200ultra
• switch, mod for sCLK-EX

 

I think this link actually allows you to order the combo, if you are US or Australia based. https://sotm-usa.com/collections/ultra-series-mods/products/sms-200ultra-audio-network-player-mods?variant=41281316428

 

[austinpop]

4 minutes ago, rickca said:

Very good point.  This focus on just the clock is like focusing only on which DAC chip is inside a DAC.  Overall design and implementation is just as important as the components, if not more so.  You can give me prime ingredients, but I still can't cook a good meal.

 

 

 

This triggered a mental image for me. Remember that show on Food Network many years ago where a celebrity chef would show up unannounced at your door, and cook you a gourmet dinner, based solely on what you happened to have laying around in your fridge and pantry? For me, that's Uptone. 

 

And then you have the gourmet meal at the Michelin rated restaurant? That's your dCS's and your Esoterics. 

 

And then the Mutec's fall in the middle.

 

Just a silly thought. 

[austinpop]

3 minutes ago, rickca said:

I can't seem to reach sotm-usa.com for the past couple of days.  Anyone else?

 

No issue here. Just went there a minute ago to find that link...

[austinpop]

2 minutes ago, rickca said:

This is too weird.  I know I've been able to access that site previously.  I cleared my browser cache and still can't get there.  I sent Kamal a note to see if he can figure it out.  It says it can't resolve the DNS address.  I usually use Chrome, and I just tried MS Edge but I had the same problem  I can reach every other website I want.  Any network gurus out there?

 

Hmm weird.

 

Fire up nslookup in a command window, and lookup sotm-USA.com? Your DNS server entries may be messed up.

 

Try hitting the site by its IP addy: 23.227.38.32.

 

We can take this to PM.

[austinpop]

1 minute ago, Superdad said:

 

Sorry Rajiv, we can't cook anything with the mustard, ramen noodles, and sad looking head of cauliflower you have in your fridge!  Let alone something that sounds like an ISO REGEN or LPS-1.   We're good, but not THAT good! 

 

But. but ... I just went to Amazon, I mean, Whole Foods!

 

Kidding aside - immense respect for what you guys produce at the price points you do!

[austinpop]

Everybody loves Ramen!

 

[austinpop]

19 minutes ago, TopQuark said:

I just want to make sure I am not interpreting this differently.  Do you mean the latter (sMS-200ultra powered by HDPlex 12V) provides significantly better SQ improvement?

 

Correct. Over the mR powered by LPS-1.

 

19 minutes ago, TopQuark said:

 

I understand from another posting that the connectors are different that may account for the difference in SQ.  Were both switches modded by SOtM?  Are you aware of any difference in how SOtM modded the switches, e.g. regulators, caps, etc?

 

At least on my Zyxel, they added the mod for sCLK-EX, replaced 2 regulators and 1 capacitor.

 

I'm not sure about the D-link.

[austinpop]

4 hours ago, lmitche said:

 

If goose-stepping matters it will be interesting to find the underlying reason. With the asynchronous, buffered component architecture used in USB audio, goose-stepping should make no difference.

 

Larry,

 

Could you define what you mean by goose-stepping? Must be an industry term. 

[austinpop]

 

54 minutes ago, rickca said:

I think Larry is talking about the synchronization of the sCLK-EX clocks due to sharing an internal reference clock.  

 

In that case, wouldn't our experiment 6, shown below be relevant? We wondered if there was a benefit to the presumed synchronization afforded by a single sCLK-EX board to all components in the chain, vs. having separate clock boards.

 

On 6/28/2017 at 1:30 AM, austinpop said:

6. Does using the same sCLK-EX board sound better/worse than separate boards?

• Compare sMS-200 mod > tX-USBultra with sMS-200ultra > tX-USBultra
• We could not hear any significant difference.

 

[austinpop]

7 minutes ago, rickca said:

In the two cases you are comparing, could you please clarify which component is hosting the sCLK-EX board?  What did you conclude from experiment 6?

 

Sure. 

• sMS-200ultra > tX-USBultra:                each component has their own independent sCLK-EX board
• modded sMS-200 > tX-USBultra:        the sCLK-EX board is on the tX, and uses two clock cables to supply the clock signals to the sMS-200.

Make sense?

 

We didn't hear any discernible difference between these configurations on Eric's system.

 

At some point, I hope to recreate this experiment on my headphone based system.

[austinpop]

1 minute ago, Bamber said:

SoTM have agreed to fit sCLK-EXs to my Wyred 4 DAC2v2 SE and a new Singxer SU-1

 

Well played, sir, well played! How did you get SOtM to agree to this?

 

I'll be following this experiment with great interest. 

[austinpop]

2 minutes ago, Bamber said:

 

I asked if they had plans for a new DDC in the near future which would include an IIS-LVDS connection and support DSD to at least 256.  My main point to them was I need a solution which will let me use a master clock on all devices in the chain.

 

They agreed to the DAC first and then agreed to the Singxer in the next email exchange.

 

I'm happy they included the Singxer even though it isn't needed in my current setup it will be a factor in planned upgrades.  It will be interesting to test further combinations in the SoTM chain.

 

Assuming you'll get sCLK-EX boards in each of the Singxer and the W4S DAC? Or share somehow?

 

While you're at it, I strongly recommend adding the switch upstream of the sMS-200ultra. I cannot overemphasize the huge bang-for-buck. Does your sMS-200ultra already have the 2 free clock taps exposed as SMB outputs on the back? If not, you will have to send it back to them to add, but well, well worth it.

[austinpop]

2 minutes ago, Bamber said:

I'm not sure how long this takes but hopefully I will be able to give you some feedback in 2 or 3 weeks.

 

I have a SMS-200ultra and am getting that fitted with the switch upgrade and a 75Ω Master Clock connection also.

 

Hah, looks like you thought of this already!

 

So your chain will look like this?

 

Switch (sCLK mod) > sMS-200ultra > tX-USBultra >SU-1 Ultra > W4S DAC Ultra

 

Mutec Ref 10 reference clock -----> sMS-200ultra

|---------------------------------------> tX-USBultra

|---------------------------------------> SU-1 Ultra

|---------------------------------------> W4S DAC Ultra

 

Way to go all in!

[austinpop]

2 hours ago, romaz said:

First of all, a big thank you to @austinpop for his hard work re-organizing this giant thread.  As I see it, Rajiv, this is now your thread, mate.  Feel free to take it where you would like.  I am happy to contribute where I can.

 

Hah - thanks, but I don't plan to moderate this thread. I do plan to continue curating an index for the first post, but that's about it. So far, I think the S/N ratio of this thread has been commendably high, TBH.

 

2 hours ago, romaz said:

I noticed Rajiv's comment on how he will be taking a break from investing further into his digital infrastructure and instead will look into putting more effort into the analog portion of his system.  Well, I have been doing the same and I just received a gigantic dose of perspective these past few days.  

 

Thanks for sharing! Most of us just take vicarious pleasure from your reports, as these are products well outside our tax brackets!  

[austinpop]

2 hours ago, romaz said:

 

At this point, I don't know just how good the clock in the sCLK-EX is nor is it that relevant to me anymore since I intend only to use this clock board as a vehicle to implement the REF10.  To be able to bring REF10 level performance to 8 components while only using 2 clock outputs still boggles my mind.

 

Roy, your phrase that I highlighted there caught my attention.

 

So if it were possible to supply two components with a reference clock, one with an "inferior" internal clock vs the other, would  the injection of the reference clock render the internal clock quality differences moot?

 

Somehow that surprises me. I would still expect the internal clock quality to carry forward. But I guess we shall see!

[austinpop]

29 minutes ago, romaz said:

Ultimately, what this means is that I have decided to abandon the AOIP route and go straight USB.  

 

I can see the direct USB "team" breaking into cheers and high-fiving!  These are awesome findings, Roy.

 

One question though - on nomenclature. Isn't the term AOIP typically used to refer to products like the Adnaco and the PS LanRover, which use IP to transport USB audio? I tend to consider the phrases "endpoint architecture" or NAA or renderer or streamer as more representative of the mR/sMS-200 approach.

 

But either way, very interesting results as usual!

[austinpop]

9 minutes ago, romaz said:

 

I'm not sure I fully understand what you're asking, Rajiv?  

 

I was just wondering what you meant by this:

 

Quote

I don't know just how good the clock in the sCLK-EX is nor is it that relevant to me anymore

 

[austinpop]

45 minutes ago, romaz said:

The comment I am hearing from several is that the IR is the "flatter" of the two but also brings richer tonal density whereas the tX-USBultra provides more atmosphere at the expense of a thinner midrange.  I could see why someone would want both.

 

Apropos of this: a quick update on @Bamber's IR/LPS-1 that he is graciously lending me. It has finally shipped  and should be here this week.

 

I do plan to burn it in for at least 100 hours before attempting a listen, so it may be a couple weeks before I'll have some results to report.

 

My intent is to try at least 3 different configurations:

1. IR instead of tX-USBultra
2. IR before tX-USBultra
3. IR after tX-USBultra

Holler if you have any other experiment of interest.

[austinpop]

6 minutes ago, ElviaCaprice said:

 

You might want to make it 6 with GI on and off the IR.

 

Good point, but just to avoid tedium, I'm not going to do all combinations. What I'll do:

1. IR instead of tX-USBultra
2. IR before tX-USBultra
3. IR after tX-USBultra
4. For whichever of 1-3 sounds best, also compare GI on vs off.

[austinpop]

14 minutes ago, romaz said:

 

It is sometimes hard to fully convey in written words the extent of how much I like something and so I'm sure there are times when people think I like something to a greater degree than I intended.  With straight USB vs the sMS-200 stack, it's not good vs horrible, it's more my personal preference as to which compromises I'm willing to live with.   Having lived with your sMS-200 Ultra stack for a while now, you can attest to how wonderful it is.  If I had chosen to remain with that stack, trust me, absolutely no one should feel sorry for me because that is one incredible trifecta, especially for the type of music I listen to, and so I would never discourage anyone from going down that route nor would I disparage anyone who is already there.  For those who feel they want more meat on the bones on their music than what this stack provides, I would also suggest adding the IR somewhere in that chain.  If there is no intention to address the motherboard, it's actually my belief that going down the sMS-200 or mR route is the better way to go.  

 

I agree with you, Roy. There's just too much either-or thinking about this out there. People should just pick the architecture and topology that makes sense for them. Your results just confirm what has always been the case - with enough attention and care, superlative SQ can be achieved with either approach.

 

14 minutes ago, romaz said:

Regarding "AOIP" or "Audio over IP", the definition I read states that this is the distribution of digital audio across an IP network.  I know that people often refer to this term with regards to such devices like RedNet and the various Ethernet protocols that such devices use but both the mR and sMS-200 are just as much Ethernet renderers as far as I'm concerned.  Devices like RedNet takes an inputs an Ethernet signal and outputs a SPDIF signal.  With the sMS-200 or mR, the output has been changed to USB and that's it.  Heck, even a basic Mac Mini or PC is an AOIP device as I see it.

 

I don't think the term is quite as broad as that, but I don't really want to or even care to argue about it.  

 

 

[austinpop]

8 minutes ago, rickca said:

So you're still going to use the bridged NIC sMS-200ultra and sCLK-EX modded switch, right?  You're just making changes at the end of your chain, is that correct?

 

Correct.

 

8 minutes ago, rickca said:

If you have time, try just listening to the ISO REGEN on its own to get some sense of its contribution.

 

Not sure I understand what you mean by "on its own." For me, that's experiment #1. But I think you're referring to direct USB right out of my music server? If so, that's not an option for me. My music server is in an adjacent room, so the distance is too great for direct USB.

 

8 minutes ago, rickca said:

Honestly, after reading romaz's change of plans today, my brain kind of hurts.  I really have been trying to follow all these experiments carefully.

 

I hear ya!!

[austinpop]

3 hours ago, romaz said:

 

A wonderfully enlightening read.  A lot of hard work on both yours and Eric's part, thanks!

 

Thanks for your kind words! It was also a lot of fun.

 

Quote

No question about convenience with 6.5-9V, especially for LPS-1 owners.  Generally, when a component consumes a fixed amount of power (watts), the advantage of going with a higher voltage supply is that less current is drawnand  to draw less current results in less noise in the ground plane.  

 

This is true in a pure Ohm's law sense, but... A DC power supply uses regulator(s) to allow a range of input voltages while providing a fixed (lower) output voltage that the component's circuit(s) actually needs. The current drawn by the component is whatever the circuit draws at this lower regulated voltage. So, say the circuit runs at 5V, but the regulators allow a voltage range from 6-14V, then the current drawn is whatever the circuit itself needs at 5V. Say that's 1A. Then the circuit draws 5W, and if the input voltage is 9V, then the regulators have to drop 4V across them, while passing through a current of 1A, which requires them to dissipate an additional 4W as heat. In that sense, the higher the difference between the input voltage, and the internal regulated voltage, the more heat the regulators dissipate.

 

I've seen @Superdad talk about this often, but I just referenced a recent post where he reiterates this point: 

Anyway, my point is that for regulated supplies, a higher input voltage does not result in any lower current, just more dissipated heat from the regulators.

 

Now, a more plausible explanation for the better SQ for the 12V Ultra could be that the regulators that SOtM used for the 12V version are sonically superior in design than what they used for the 6.5-9V version.

 

Still, all in all, we didn't find a noticeable difference here.

 

 

Quote

 

 

... there is probably more to why the SU-1 sounds good then just the clock.  For starters, the XMOS USB interface used by the dX is quite old (2nd generation) while the SU-1 uses the latest version.  I think this is a big deal and it sounds as if the dX is in need of a freshening up.

 

Yes, Eric and I had the same thought about the older XMOS platform.

 

Quote

I have posted this same impression regarding the iSO-CAT6 isolator several times on this thread already and so this observation comes at no surprise.  There is definitely some improvement in terms of alleviating a certain harshness but the improvement is small.  I use it because I already own it and that's it.  

 

Yes, this continues to be my experience with Ethernet isolators too. I know @lmitche and some others have found dramatic improvements with this class of device (EmoSystems, Etalon), which is puzzling. Like you, in my experience, neither the iSO-Cat6 nor my earlier Baaske made much of an impact at all.

 

Quote

 When I first reported on this, the main impact of the dCBL-CAT7s were within the direct connection pathway.   ... Within the "direct path" between the server and the sMS-200ultra, I did find the impact significant and worthwhile. ... My only complaint about this is how thick and unwieldy these CAT7 cables are.  Another reason I chose to go straight USB.

 

To be clear, we also used the dCBL7 in the direct path only, not upstream of the Roon server. Did I give that impression? If so, apologies.

 

I did find the dCBL7 to improve the SQ. This is honestly the first time I ever heard an Ethernet cable to sound better. But I didn't think it was $500 better. Certainly not anywhere near as significant as the addition of the switch. So in this regard, we clearly hear different magnitudes to these effects. Very likely, your system is far more resolving.

 

Eric will gladly lend me these cables to try in my setup, so I will reassess the effects of the iSO-Cat6 and the dCBL7 in my setup. 

 

Finally, I heartily concur about the unwieldiness and stiffness of these cables!

[austinpop]

3 hours ago, rickca said:

For those using SOtM-modded ethernet switch and sMS-200ultra, what if you added the JCAT NET Card Femto? 

 

TBH, this sounds very interesting, but I don't know what to expect. When we added a 2nd switch in the direct connection - to add another sCLK-EX reclocking - the incremental SQ improvement was small.

 

So is that a good proxy for what to expect with the JCAT femto card? Or will the improvement be greater/worse? We just don't know.

 

Anybody going to to try one?!

[austinpop]

5 hours ago, romaz said:

Here's what SBooster has to say about using a low voltage vs high voltage PSU for the sMS-200 specifically with respect to SQ.  Seems like a difference of opinion.

 

Agreed. In practice, it seems to be a moot point, vis a vis SQ of the sMS-200ultra. 

 

May did send me another intersting explanation.

 

"When using 7V to the sCLK-EX,  one of constant voltage circuit is not in used, but when using 9V or 12V to the unit, the constant voltage circuit is used to block noise, so it effect sound. so we recommend using many users that there are sound differences when using 12V and 7V. But we always inform together that the sound or the sound differences can be made by the personal taste and by what system is used. So 12V power can’t be always the better, it could be differ from what system is or the personal taste."

[austinpop]

2 hours ago, lmitche said:

Rajiv,

 

The Adnaco doesn't use IP at all.  It is extending PCIE serial connections over fiber and therefore uses pcie bridging and bus protocols.

 

Ah I see. I didn't realize that it was extending PCIe. I assumed it was extending USB. My bad.